Cutting Edge Harvard Case Solution & Analysis

Cutting Edge Case Solution

Question: 2c.

The five days’ moving average method has the specialty of using multiple data points and the recent history to forecast the new value. It is explained as the average of data for the n period of time and Harry used n as 5,which means that the data of last 5 days was taken for forecasting. Further, it is also argued against the method that this method may ignore the changing environment because it allocates same weight to all the selected values. And in the accuracy chart; moving average is listed at third,with MAD value of 218 as compared to the other forecasting methods.

Question: 2d.

As exponential smoothing method is bit same as moving average method, the only change is that it gives weight to the recent observation and allocates a smaller weight to the next progression towards the older data point.  It has alpha in its formula and this alpha defines the stability of the process.Whereas,Harry used 0.1 alpha because he stated that the conditions will remain stable and he will get the MAD of 249, and it is ranked forth according to its data accuracy.

Question: 2e.

Increasing the alpha value represents that the recent vales are given higher weights as compared to previous values in the time series data set. The variability in the output is high, because the response towards the conditions is fast, which creates high variability in the results, and in the accuracy table; this method is ranked on first position.

Question: 2f.

After analyzing the aforementioned forecast methods; we can clearly see that if MAD is low then it is accurate forecasting. So, we can say that the mean values should not deviate from the actual,and that is to reduce the MAD value.

Question: 3a.

 

s.no Year Month Volume Forecasted
1 2014 Jan 24,015
2 2014 Feb 25,203 24,015
3 2014 Mar 23,589 24,609
4 2014 Apr 26,704 24,099
5 2014 May 28,120 25,402
6 2014 Jun 26,321 26,761
7 2014 Jul 27,021 26,541
8 2014 Aug 25,981 26,781
9 2014 Sep 26,456 26,381
10 2014 Oct 27,120 26,418
11 2014 Nov 26,954 26,769
12 2014 Dec 27,321 26,862
13 2015 Jan 26,456 27,091
14 2015 Feb 27,450 26,774
15 2015 Mar 31,580 27,112
16 2015 Apr 33,124 29,346
17 2015 May 32,432 31,235
18 2015 Jun 31,901 31,833
19 2015 July 31,867

Call Volume Forecast for July 2015 exponential smoothing is 31867.

Question: 3b.

 

Year Month Volume  Count Forecast
2014 Jan 24,015 62,120 23967.53
2014 Feb 25,203 62,152 24401.06
2014 Mar 23,589 62,138 24834.59
2014 Apr 26,704 70,343 25268.12
2014 May 28,120 69,120 25701.65
2014 Jun 26,321 68,967 26135.18
2014 Jul 27,021 67,956 26568.71
2014 Aug 25,981 65,342 27002.24
2014 Sep 26,456 65,380 27435.77
2014 Oct 27,120 65,432 27869.3
2014 Nov 26,954 65,423 28302.83
2014 Dec 27,321 65,650 28736.36
2015 Jan 26,456 65,620 29169.89
2015 Feb 27,450 65,610 29603.42
2015 Mar 31,580 77,231 30036.95
2015 Apr 33,124 75,201 30470.48
2015 May 32,432 74,978 30904.01
2015 Jun 31,901 75,012 31337.54
2015 July 31,771

 

The forecast for July 2015 is 31771.

Question: 3c.

 

s.no Year Month Volume reg forecast Forecasted MAD Ab. Error Ab. Error
1 2014 Jan 24,015 23,968 24000 15 47 47.47
2 2014 Feb 25,203 24,401 24,015 1,188 802 801.94
3 2014 Mar 23,589 24,835 24,609 -1,020 -1,246 1245.59
4 2014 Apr 26,704 25,268 24,099 2,605 1,436 1435.88
5 2014 May 28,120 25,702 25,402 2,719 2,418 2418.35
6 2014 Jun 26,321 26,135 26,761 -440 186 185.82
7 2014 Jul 27,021 26,569 26,541 480 452 452.29
8 2014 Aug 25,981 27,002 26,781 -800 -1,021 1021.24
9 2014 Sep 26,456 27,436 26,381 75 -980 979.77
10 2014 Oct 27,120 27,869 26,418 702 -749 749.3
11 2014 Nov 26,954 28,303 26,769 185 -1,349 1348.83
12 2014 Dec 27,321 28,736 26,862 459 -1,415 1415.36
13 2015 Jan 26,456 29,170 27,091 -635 -2,714 2713.89
14 2015 Feb 27,450 29,603 26,774 676 -2,153 2153.42
15 2015 Mar 31,580 30,037 27,112 4,468 1,543 1543.05
16 2015 Apr 33,124 30,470 29,346 3,778 2,654 2653.52
17 2015 May 32,432 30,904 31,235 1,197 1,528 1527.99
18 2015 Jun 31,901 31,338 31,833 68 563 563.46
19 2015 July 31,867 873 Abs. error 1292.065

The mean absolute deviation of exponential smoothing model is 873 when alpha is 0.5 and absolute error is 1292.065 so the difference between both values is due to the user counts in both data. As this take count and the previous do not take any count..............................

Cutting Edge Case Solution

This is just a sample partial case solution. Please place the order on the website to order your own originally done case solution.

Share This

SALE SALE

Save Up To

30%

IN ONLINE CASE STUDY

FOR FREE CASES AND PROJECTS INCLUDING EXCITING DEALS PLEASE REGISTER YOURSELF !!

Register now and save up to 30%.